II. Warp Maneuvering
III. Warp Combat and Strafing
A. Strafing Events
B. Running the Numbers
C. Troubleshooting: The Utility of Warp Strafing
1. Eye of the Needle: Targeting at Warp
2. Attack Intensity
3. Defense and Return Fire
c. "Way of the Warrior"
d. "A Call to Arms"
e. "Tears of the Prophets"
f. Warp-Impaired Vessels
See Also: Objections Page
The concept of faster-than-light travel having strategic utility is readily comprehensible. The strategic utility of rapid transit of personnel and assets has been a fixture of military thought since the roads of Rome, and long before. However, the finer tactical utility of Star Trek's warp drive is often overlooked.
Whereas Star Wars hyperspace is an alternate dimension wherein vessels are virtually immune from realspace attack and incapable of attacking objects in realspace, warp drive involves the use of subspace fields to propel a vessel beyond lightspeed within the realspace dimension we're used to. This means a Federation starship can affect and be affected by opponents who are at sub- or supralight speeds.
In other words, one can readily imagine the possibility of a sublight target being blasted by a vessel engaged in a warp flyby . . . such an incident was shown as occurring in the 2260's, in "Journey to Babel"[TOS] wherein the Enterprise warp drive was disabled. This attack pattern has been referred to as "warp strafing", calling to mind strafing runs of fighter planes in the early 20th Century.
By analogy, if sublight speed is akin to ground travel, then warp would be like air travel. And just as in the 20th Century aerial dominance allowed one to attack ground targets with virtual impunity, so too did the Orion attack of "Journey to Babel" show the same sort of "subspace dominance". It was only when Kirk tricked the Orions into "landing" (figuratively speaking) by dropping out of warp that he was able to return fire effectively. Indeed, one can readily imagine a situation where a vessel engaged in a warp flyby could dominate an entire slower-than-light fleet armed with slower-than-light weapons, just as ground targets not equipped with weapons intended for use against aircraft can be dominated by attack planes.
Of course, with most of the Star Trek races every vessel is, to extend the analogy, a fusion of airplane and tank, capable of operations both at slower-than-light and faster-than-light speeds with nearly instant conversion between the two. One can imagine the futility of attempting to strafe based on one's aerial dominance if, all of the sudden, opposing tanks all took to the air firing air-to-air missiles.
Thus, the concept of warp strafing brings with it many questions and a few complications. In the below, we'll look at the required warp maneuverability involved and, further, examine the utility of warp strafing as a combat maneuver.
In the episode "Fury"[VOY6], we get a glimpse into a (slightly rewritten) past. It is set during Voyager's first season, and the subplot involves the organ-harvesting Vidiians trying to force Voyager toward a sort of 'warp cul-de-sac' so the ship can be captured (equivalent to forcing an aircraft to land). This warp cul-de-sac is an area of space littered with "subspace vacuoles", covering seventy percent of the region. Janeway decides to traverse the region anyway, but a special technique has to be used:
Janeway: "Tom, what's the first thing they teach you about maneuvering at warp?"
Tom: "'Faster than light, no left or right.' When possible, maintain a linear trajectory. Course corrections could fracture the hull."
Janeway: "Exactly. We'd have to drop to impulse every time we made a course change."
Voyager ends up engaging in a number of warp mini-jumps, dropping to impulse to turn before jumping again. The region of subspace vacuoles required 200 such course changes.
That event, and the quote above, have frequently been misunderstood to imply that a warping starship must maintain a straight-line trajectory and cannot turn. However, that would be inaccurate. The statement above is the first thing taught . . . not the last. Further, the edict is specified as being the preferred option "when possible". A linear trajectory can thus be said to be preferred, but not required. In the case of "Fury", it must logically have been the extreme maneuvering required that called for the stop-and-go warp flight.
The above suggestion is far from a baseless claim. After all, numerous vessels shown across 200+ years of Trek history have maneuvered without incident:
Enterprise's "Shockwave, Pt. II" shows Suliban vessels surrounding the ship and firing on her at warp, maneuvering effortlessly. When Enterprise's hull plating begins to fail on the port side, T'Pol orders a ten-degree turn to the starboard. We see this occur while the ship remains at warp speeds.
In "Future Tense"[ENT2], the Tholians demonstrate the ability to turn at warp by engaging in a 30+ degree upward turn after harassing Enterprise.
In "Canamar"[ENT2], a lowly prison barge banks to port while at warp, as per an ordered course change.
"Divergence"[ENT4] features the Earth starship Columbia pursuing Enterprise (whose warp engines are in a runaway state), catching up to her, rolling 180 degrees, and maneuvering into position directly beneath.
Then there's "Operation: Annihilate"[TOS1], in which the Enterprise pursues a Denevan ship towards the Denevan sun at a stated speed of warp eight. The Enterprise doesn't make it in time, however, and after the Denevan ship burns up, Kirk orders "one-hundred eighty degrees, hard about". We see the viewscreen indicate that the ship turned quickly to port, and it is only several seconds later that Kirk orders a reduction to "sub-warp speed".
In "Elaan of Troyius"[TOS3], the Enterprise warp drive has been sabotaged, and then a battle begins against a warp seven Klingon ship engaged in strafing. The bridge conversation implies strongly that the Enterprise would normally be maneuvering at warp in such a battle, since the impulse maneuvers are "sluggish" and make the ship "wallow like a garbage scow against a warp-driven starship." This notion is confirmed when, the engines repaired, Kirk orders a *pivot* at warp two, and the ship fires. The battle is won.
In "The Deadly Years"[TOS2], a Starfleet commodore has foolishly taken the Enterprise into the Romulan Neutral Zone, where she is being pursued and attacked by "a maximum of ten" Romulan Bird-of-Prey class ships. The Enterprise sailed along at warp five. When Kirk resumes command, he employs a variation of the Corbomite Maneuver, and as the Romulans back off, he has the Enterprise engage in a simultaneous course change and acceleration to warp eight. We see the ship turn with extreme rapidity.
In "The Changeling"[TOS2], the battle with Nomad causes the loss of "warp maneuvering power" due to the drain on the engines (via the shields). Naturally, if they lost it, then they previously had it.
From "Let That Be Your Last Battlefield"[TOS3], we have this gem:
Kirk : "Mister Spock, is this ship headed for Ariannus?"
Spock : "Negative captain. The Enterprise is now moving in a circular course."
Scotty : "And at warp 10 we're going nowhere mighty fast."
Circling, and at warp 10 no less!
In Star Trek IV: The Voyage Home, the Klingon Bird-of-Prey does some serious maneuvering around the sun, and it cannot be explained as the ship simply `being maneuvered' by gravity.
In "Conspiracy"[TNG1], the Enterprise-D, en route to Pacifica at warp eight, detects a disturbance in a nearby sector (caused by the self-annihilation of the USS Horatio). Picard decides to investigate, telling Geordi to "maintain speed, alter course accordingly." The next scene shows the Enterprise-D at warp speed, logically the very same warp eight at which she started.
In "The Wounded"[TNG4], the Enterprise-D is flying in close formation with the Nebula Class starship Phoenix, escorting that ship back to Starfleet for having attacked the Cardassians without orders. The captain of the Phoenix, however, isn't done being naughty, and changes course. We can actually see what Picard see, as he observes the Phoenix on the viewscreen . . . it engages in a slow banking turn to starboard, at warp, the 'warp stars' streaking by.
Also, by default, starships that can separate at warp (such as the Galaxy Class or the Prometheus Class from "Message in a Bottle"[VOY4]) must maneuver at warp at least a bit in order to achieve separation (the Prometheus separated widely at warp). "Encounter at Farpoint[TNG1] shows us the separation of a Galaxy Class ship at warp, with the stardrive section cutting 180 degrees hard to starboard (see also Objection 1). This turn-on-a-dime is almost Kirk's "warp pivot" move from "Elaan of Troyius"[TOS3].
There are many other scenes which involve starships at warp changing course, though the actual event is not seen. One example is "The Ultimate Computer"[TOS2], when the M-5 computer changes course to go after a sublight freighter, performing warp strafing moments later. None of these scenes ever involve anything like the phrases "drop to impulse so we can change course", or even anything like the concept. Janeway, in "Human Error"[VOY7], actually orders evasive maneuvers at warp. We see some idea of what these would look like when Odo, piloting the venerable old runabout Rio Grande, evades, jinks, and dodges around (and above) a Jem'Hadar attack fighter in "Treachery, Faith, and the Great River"[DS9-7], finally destroying it.
The Voyager episode "Fury" is inconsistent with these examples, when the statements are misunderstood to suggest that warp maneuvering is impossible. While maintaining a linear course might be preferred, there are numerous examples to show that any hull stress must be marginal, even on lower-tech transport ships. After all, an old Constitution Class starship easily survived a pivot at warp speeds, which would presumably involve all sorts of ugly forces that would try to tear the ship into tiny fragments of its former self. And, a Galaxy Class stardrive section survived a high-warp turn of 180 degrees. Therefore, simple sweeping slow turns simply aren't going to be that bad.
One unusual thing about warp maneuvering as observed is that it isn't as violent a procedure as we might expect. The images above all show scenes of a vessel seeming to engage in banking turns . . . but at the velocities involved the ships ought to shoot out of sight within the first millisecond if they're heading the way their noses are pointing.
As example, let's imagine two ships travelling in tight formation at twice the speed of light. One vessel executes an instant course change of one-tenth of one degree. At a velocity of just under 599,585 kilometers per second, in one second after the course change the vessels will be over 1000 kilometers apart (1,046.47, to be more exact). That's just for one-tenth of one degree . . . not even an entire degree, and certainly not the 30 degree change we saw the Tholian ship engage in in the Enterprise shots above. In that case a huge change of apparent direction produced only a tiny change in actual course, measurable in perhaps the hundredths of a degree, if that.
Clearly, then, warp maneuverability isn't quite as grand as it could be in some cases. The Tholian example and the Phoenix example are perhaps the worst, and although there are examples that are vastly superior (e.g. Deneva, the pivot, the Galaxy stardrive turn, etc.), few seem to equal the instantaneous course change result we calculated a moment ago.
We may therefore presume that there is a bit of subspace inertia or some similar effect that largely prevents warp turns from being quite as impressive as we might imagine. Many vessels engaged in a warp turn might seem to be almost "power sliding" their way through the turn to an outside omniscient observer. It isn't clear how this inertia is overcome in cases that don't seem to feature it, though this "warp drifting" should not be ignored because it does not always appear.
Combat between two faster-than-light vessels is nothing new, be it with phasers or torpedoes. We've seen it occur as often as we've seen ships maneuvering at warp, from the Suliban assault on the NX-Enterprise in "Shockwave, Pt. II"[ENT2], to the aft torpedo firing from "Encounter at Farpoint"[TNG1], to the beam weapon combat in "Treachery..."[DS9-7] and "Message in a Bottle"[VOY4]. Last but not least, there's the forward-firing torpedo shot from Klingon ships pursuing the Excelsior in "Flashback"[VOY3]. The only implication that FTL combat might cause trouble is from "Fallen Hero"[ENT1], where Reed says that has not yet properly adapted the phase cannons to fire without disrupting the warp engines.
On the other hand, combat between a warp-driven starship and a sublight vessel or target (a.k.a. "warp strafing") is a far more rare occurrence. After all, if you're in the midst of impulse combat and your opponent goes to warp to achieve the strafing advantage, it's hardly inconceivable that you'd zip to warp, likewise. In the case of targets restricted to sublight, however, warp strafing could provide a devastating advantage, in situations where it is compatible with one's goals.
We've seen a few examples of warp strafing in the canon.
In "Elaan of Troyius"[TOS3], the Enterprise's warp drive has been sabotaged. A Klingon battlecruiser makes several warp and impulse passes against the crippled Enterprise, scoring hits with every disruptor shot. Scotty performs a miracle, and Kirk waits for the right moment to surprise the Klingons with his newly-restored warp capability. After the final pass of the battlecruiser (which occurs at sublight), the Enterprise leaps into a warp two pivot, firing photon torpedoes against the battlecruiser. The Klingon ship is "badly damaged", and limps away. (The Klingon warp strafes are undeniable, but there is a possibility that the Klingon ship had jumped to warp after strafing the Enterprise that last time (as opposed to the Enterprise engaging in a return warp strafe), though no mention of this is made on-screen.)
In "The Ultimate Computer"[TOS2], the Enterprise, under the control of the M-5 computer, attacks the sub-light freighter Woden. The Woden is torpedoed by the Enterprise, which was moving at a stated speed of warp four.
Note: Some claim the vessel was warp-capable. However, there is no known Starfleet or Earth design of that era which lacked brutally obvious warp nacelles (even the Aurora had them in "Way to Eden"[TOS3] in spite of all aesthetics, as did the Y-Class freighter seen in "Fortunate Son"[ENT1], the Conestoga from "Terra Nova"[ENT1], etc.). Further, the vessel seen on screen is of the old DY configuration, a la Khan's sublight sleeper ship. Further, she is designated as "slow-moving", which would make little sense if understood as anything but sublight, given that the Enterprise's speed at the time the comment was made was somewhere below warp three. However, the remastering of ST:TOR might change this.
Some have claimed that photon torpedoes cannot be used for warp strafing, but this notion is absurd. Besides the fact that we've seen it done in "The Ultimate Computer" and almost certainly in "Elaan of Troyius", there's plenty of supporting evidence:
"Elaan of Troyius"[TOS3], et al. - late 2260's:
Beam weapons may be used to attack sublight targets from a warp-driven ship. Those are unguided, point-and-shoot weapons . . . why should a guided torpedo suddenly be unable to do the same thing?
If they can point the ship and her phasers in the direction of the enemy, and achieve the zillionths-of-a-degree targeting capability necessary to hit the enemy from the ship with a beam, then there is no reason a guided torpedo fired in a similar fashion can't hit the target, especially given that the torp's odds will improve as she gets closer (and thus the target becomes larger and easier to hit).
"Journey to Babel"[TOS2] - late 2260's:
Kirk evidently hoped to be able to hit the Orion vessel with the phasers or torpedoes he was firing as the two ships passed. The Orion ship was moving at a stated speed of warp 8, though she was capable of faster speed. The Enterprise's speed is unknown. This would be a sort of reverse-strafe example.
"Maneuvers"[VOY2] - early 2372:
Voyager plans to rescue Chakotay with a dangerous transporter technique: transport of a stationary target by a vessel passing at warp speed. Torres says she has done it before, presumably when part of the Maquis (which suggests, given their limited and out-of-date tech, that it isn't hard).
(Logically, accurately scanning and dematerializing a two-meter tall person while passing by at warp should be far, far harder than pinging a dozens- or hundreds-of-meters long starship with a torpedo.)
"Balance of Terror"[TOS1] - late 2260's:
When running from the warp-speed Romulan plasma weapon, Kirk did not order a drop to sublight. If a warp weapon cannot hit a sublight target, this would have been an obvious move.
"The Changeling"[TOS2] - late 2260's:
The Enterprise does battle with Nomad, a stationary one-meter target. Scotty diverts partial warp power to the shields, but the ship maintains "warp maneuvering power". This is lost prior to the firing of the torpedo (thereby making this *not* a direct example of torpedo strafing), but the fact that he chose to arm them while at warp (as opposed to the phasers) is quite telling. When the Enterprise does fire, she shoots from a stated range of 90,000 kilometers.
(Note: There were a few seconds between this statement and the torpedo firing, during which point the Enterprise loses warp maneuvering power. However, another shot by Nomad (whose bolts have a stated speed of warp 15) takes a couple of seconds to reach the ship, meaning that 90,000 kilometers serves as a ridiculous lower limit.)
Let's stop for a moment and consider the example from "The Changeling". The Enterprise, at 90,000 kilometers minimum from Nomad, achieves a direct hit. While that was an impulse shot, we may presume that the targeting capability demonstrated might also apply to warp strafing. So, let's ponder:
From the ship, and assuming Nomad were a full one meter sphere instead of a slender one-meter tube, that's still a target merely 0.0000006366 degrees (0.00229 arc-seconds) in width:
.5 A / D = tan (.5 (theta)) .5 1m / 90,000,000m = tan (.5 (theta)) .5m / 90,000,000m = tan (.5 (theta)) 5.5555e-9 = tan (.5 (theta)) tan (.5 theta) = 0.00000000555 .5 theta = tan-1 (.00000000555) .5 theta = 3.18309e-7 theta = 6.3661e-7 theta = 0.0000006366 degrees
Note well, a 1.6 kilometer Imperial Star Destroyer will appear to be .00229 arc-seconds in width (broadside) at 144,000,000 kilometers. Assuming that a torpedo were fired at that range and was moving at twice the speed of light, this would give the torpedo 240 seconds to home in on the target. At 200c, the torpedo fired from that range would have 2.4 seconds to home in on the target. Given that we have seen torpedoes travel, by one approximation, at least 600,000,000 kilometers over the course of a couple of seconds at warp and successfully strike the target, this doesn't present a problem.
What does present a problem, though, is that in "Journey to Babel" and "Elaan of Troyius" the Enterprise is warp-strafed by vessels firing beam weapons at high warp . . . better than warp seven from the Klingons, and warp eight for the Orion vessel.
Even if we assume a speed of only 1000c for these examples, then we're talking about 299,792,458km/s as the speed involved. Assuming a ship has to fire at least one second away from the target, then that would suggest a firing range of about 300,000,000 kilometers.
This tells us that the .00229 arc-second standard is actually not tight enough, since one wouldn't be able to hit the broad side of a kilometer-wide barn at that speed. An actual shot from 300,000,000 kilometers hitting the 289m Enterprise would require the ability to strike a target with an apparent width of only:
.5 A / D = tan (.5 (theta)) .5 289m / 300,000,000,000m = tan (.5 (theta)) 4.8166666667e-10 = tan (.5 (theta)) tan (.5 theta) = 0.00000000048166666667 .5 theta = tan-1 (0.00000000048166666667) .5 theta = 2.75975e-8 theta = 5.51949e-8 theta = 0.0000000551949 degrees
0.0000000551949 degrees is 0.000198702 arc-seconds, or just a smidgen below 2 milli-arc-seconds. To give you a sense of how extreme that figure is, consider that with such accuracy you could shoot a man in the eye at a range of about 230 kilometers. That's roughly like shooting someone in the eye across the English Channel, or from beach to beach across the state of Florida.
And that's not with a guided, homing torpedo . . . those were beam weapons. Indeed, the fact that in "Maneuvers" the ship was able to target a stationary person's molecules for beaming while flying past at warp tells us that the concept of the homing torpedo above may not be needed. After all, only either the torpedo or the ship has to have the targeting capability to hit such a microtarget . . . the weapons don't need the capability if the ship has it, though it would be helpful. The reverse is also true. It would have to be proven that neither the ship nor the weapon is capable of that level of targeting, but this is impossible . . . we know at least one is (i.e. the transporter example), and we know from the prior examples of torpedo warp strafing that one or the other has been the case for a long time, and is sufficient.
We don't know if the torpedoes already seen to be used for warp strafing were fired dumb, set along a pre-programmed course, given telemetry updates from the ship, or guided themselves in. Given the targeting ability already displayed in sublight and FTL, however, one of the latter two is most likely. And, given Worf's comments about the torpedo's onboard guidance system in "Genesis"[TNG7] and Timicin's solar guidance program in "Half a Life"[TNG4], it's likely the last option.
Given that we've seen warp strafing, the question becomes why we haven't seen it at other times when it might've been opportune.
Using the two different arc-second widths given above as a standard, let us consider a range of possible sizes and ranges at which that apparent width would be of value:
(Note: If you wish to calculate a specific value yourself using the .00229 arc-second standard, all you need to do is halve your sphere diameter then divide the result by 5.5555E-9. Similarly, the other value can be found by halving and dividing by 4.8166666667E-10.)
Given that a speed of 2c would mean that a torpedo was travelling at 599,584.9 kilometers every second, one can see that targeting a 5m wide spot on an enemy vessel could be awfully tricky. If the firing ship were also travelling at about 2c it would have less than one second to fire and evade by the .00229arcsec standard. And, at the higher velocities seen in TOS warp strafing, the situation is worse . . . a vessel travelling at 300,000,000km/s would have one-sixtieth of a second to fire and then evade.
Judging by the chart and the assumption that a vessel needs at least a second to fire and then evade, the problem becomes clear pretty quickly. Smaller or very specific targets become almost impossible to hit at high speeds, and even at low speeds the chances are slimmer.
Warp strafing can thus be understood to be a club, not a scalpel.
Warp-strafing presents certain advantages, but one of those is, ironically, not speed. A warp-strafe attack will be less intense a bombardment than an impulse battle, simply because the attacking vessel will only be able to fire once or maybe twice per pass. While the time between passes will depend on many factors, certainly it must be many seconds before a vessel at warp could swing around for another shot. Compare this to a battle at impulse, where the rate of fire is strictly dependent on weapon arcs and recharge rate, and it's easy to see how a warp strafing attack might be a less efficient attack method in some circumstances.
There's also the fact that starship battles in the 24th Century are also often point-blank affairs. The advantages this confers are lost in a warp strafing situation.
Above, we used an example of both a warp-fired torpedo and a warp-strafing ship travelling at 2c. There was good reason for this. The torpedo will not have a significantly greater speed than the ship . . . weapons fired from vessels at warp do not appear to gain more than perhaps a fraction of lightspeed in relative velocity, and certainly not additional warp factors.
Coupled with the problem of warp drifting mentioned above, then a warp-strafing attacker has a problem.
For starters, a warp-strafing vessel won't present too great a surprise when it comes to attack vectors. If a warp-strafing ship is approaching from dead astern, it isn't like its fire is going to be able to hit your forward shields . . . you can count on a hit to your aft sections and prepare accordingly.
But more specifically, the approach vector can't be too much of a surprise, either. With warp drifting often resulting in course changes equivalent to less than a tenth of a degree, some pretty good guesses can be made as to where the strafing ship will pass. This was undoubtedly the idea behind Kirk's attempts to fire on the Orion vessel in "Journey to Babel", though apparently the Orion ship was strafing at such a high velocity (with a commensurately-high distance gained via its evasion maneuver) that hits simply could not be scored.
This may suggest that low-warp-velocity strafing just isn't useful even in the 23rd Century against targets which are of 23rd Century Federation tech level but which, for whatever reason, are restricted to sub-warp speed. A warp-strafing attacker has to be screaming along at high warp in order to avoid receiving hits.
Further, although the maximum range for a warp-strafe shot against a sublight target is not known, there will undoubtedly be a range limitation, one which curtails the possible evasion distances and vectors.
Fixed fortifications would seem to be a prime target for warp strafing, be they in the form of planets or starbases. Indeed, examples from Deep Space Nine battles have been the primary evidence used in attempted counterclaims.
Although the Cardassians left their orbital ore processing and administrative station Terek Nor largely in ruins upon their pull-out from Bajor, Starfleet and the Bajorans were able to get it at least partially operational in short order. When the wormhole was discovered, the station, now known as Deep Space 9, was moved to the mouth of the wormhole. When Dukat's Galor Class ship enters the wormhole and disappears, three additional Cardassian Galors arrive and threaten the station, demanding its surrender in retaliation for the missing and presumed-destroyed vessel.
The battle is short-lived. The Cardassian Gul notes that the station could not defend itself against a single Cardassian warship at that point, but he of course had three. Though Kira tried to feign greater weaponry (and, in a cautious move, Jasad called for reinforcements), Jasad had little doubt that he could defeat the station. DS9, after all, could only bring partial shielding up. And Jasad was correct . . . with the exception of a single high-yield phaser hit to one of his vessels, Jasad wore down DS9's defenses almost effortlessly.
In short, then, this event could've involved warp strafing, but there was simply no need of it.
In an alternate timeline, a Romulan Warbird suddenly decloaks and attacks Deep Space 9. Unprepared for the assault, the station is quickly mortally wounded, and the station's survivors escape shortly before the station explodes.
As is the case with "Emissary", the station's weak defenses in "Visionary" mean that warp strafing would've been entirely superfluous.
In this episode, the Klingon Empire has invaded Cardassia under the belief that its recent coup was not the civilian uprising commonly believed, but in fact a Dominion plot. The Federation does not have any evidence to support this and the Klingons provide none, so as a result the Empire and Federation find themselves at odds. The Federation refuses to participate in and condemns the invasion. In response, the Klingons withdraw from the Khitomer Accords, the major treaty of peace and alliance between the two powers.
Soon enough, the Cardassian Union, already in a state of disarray due to the earlier loss of its "Obsidian Order" secret police and now the civilian takeover, loses a great deal of territory quickly to the Klingons, who are invading with approximately one-third of their military. Commander Sisko, in communication with the military attache to the new Cardassian civilian government, hatches and executes a plan to spirit away the civilian ruling body, the Detapa Council, from Cardassia to DS9 in the hopes of avoiding their capture and execution by Klingon forces. DS9, with its defenses recently upgraded, is intended to serve as a protective base.
Thus we have the setup for the episode's battle, wherein a detachment of the Klingon fleet comes to DS9 demanding the handover of the Detapa Council. Sisko refuses, and the Klingon fleet attacks at close range. The battle is fought entirely at slower-than-light velocities, and features boarding action by the Klingons.
While some would claim this as a disproof of warp strafing, though, such a view is not accurate. Even if we assume that DS9 had no defense against warp strafing, the Klingons also had no cause to use it. Their goal was to bring down the station's shields and board DS9 via transporter, intending at least to take the Detapa Council and possibly to take the station altogether.
This goal is largely incompatible with warp strafing, especially in regards to use of the transporter. Certainly the fleet could've warped about while working on the station's shields, but the response time to boarding the station would've been lessened . . . as seen, bringing down the shields only lasted for a short time before they were raised, at least temporarily. Splitting up the fleet to have some vessels warp strafing while others remained within or near the 40,000 kilometer transporter range would also not work, since the sublight vessels (loaded with boarding parties) would then receive all of the station's fire. (Weapon ranges for Cardassian and Starfleet vessels is known to exceed 200,000 kilometers.)
In short, this situation would serve as an exception proving the rule.
With the Dominion re-arming of the Cardassian Union forcing Starfleet's hand, Sisko mines the entrance to the wormhole. The Dominion sends a fleet to attack the station, a fleet whose purpose is made clear by the Vorta Weyoun: "As I see it, you have two choices: either remove the mines, or we will take this station from you and remove them ourselves." Further, the Cardassians (especially Gul Dukat) were very interested in the retaking of DS9.
With capture of the station the primary goal, the situation is akin to that of "Way of the Warrior". The Dominion forces did not seek to severely damage the station . . . the plan was to force the station's surrender, taking it and then taking the minefield down.
Dukat's battle strategy involved a primary battle group plus vessels in reserve waiting outside the Bajor system.
Note well that up until this battle, Dominion weaponry had never been impeded by Federation shielding. The attack plan was thus undoubtedly based on the seemingly-weak structure of the station, a structure that cannot withstand an assault by single-digit numbers of warships as seen in the earlier-season examples. Dukat thus believed he could take the station before it could do significant damage to his fleet with its greatly enhanced weapons loadout.
Dukat's battle strategy for the primary battle group involved impulse-speed passes of groups of the heavy ships, seemingly with more constant harassment by other vessels. The initial pass resulted in something astonishing to Weyoun and the Cardassians . . . the station's shields held. Thus Dukat orders a second pass with all vessels targeting a certain set of shield coordinates on the docking ring in an explicit attempt to penetrate the shields. The attempt is at least partially successful given that main power to the shields is lost and given the massive damage done to that section of the docking ring, but Worf is able to restore the shields with auxiliary power.
A continued assault at that point might've brought the shields down altogether and allowed for the anticipated Jem'Hadar boarding operation, but when the minefield is brought online the Dominion forces regroup and the reserves are called in to the system for a "final assault".
However, given that the Federation strategy was to get the minefield up and give up the station in favor of a large Federation-Klingon fleet attack on Dominion shipyards in Cardassian space, Sisko and the rest of the Federation personnel piled into the Defiant and Rotarran and departed.
Of course, warp strafing could've been used instead of the impulse passes actually employed. Per Weyoun at the end of the episode, the battle cost the Dominion-Cardassian fleet over fifty ships. So why didn't Dukat use it? The answer would seem to be twofold. First, as with the Klingon attack from "Way of the Warrior" the boarding of the station seemed to be desired, though the Federation withdrawal nullified that plan. Second, the Dominion fleet was taken by surprise with the Federation shielding, and the response was to target a very specific area of the shields with fire from multiple axes . . . something that might not have been possible in a warp strafing situation.
The Federation-Klingon-Romulan alliance executes a fleet attack against the Chin'toka system, a system with a main double-planet system surrounded by a defense grid composed of hundreds of a new type of Cardassian automated orbital weapon platform, grouped in clusters. The defense grid is still offline as the fleet approaches, but as the fleet begins dispatching the platforms one-by-one the grid is finally ready and brought online. The fleet is thus caught within the grid when the platforms begin firing.
The Alliance fleet takes massive damage, and in the meantime quickly learns that the shields of the platforms cannot be penetrated. It is soon realized that the platforms share a common power generator on an asteroidal moon, but an attack directed there is also ineffective. A technobabble trick makes the platforms fire on their own power generator, destroying it and thereby disabling themselves.
Warp strafing would not have been thought necessary up until the point where the platforms became active. At that time the fleet should not have been stuck . . . the platform density was high in the clusters, but mere hundreds of platforms across an area containing two planets could not have resulted in a lack of warp exit vectors.
The problem would thus seem to be the small size of the platforms mixed with the knowledge that only a very intense attack could hope to penetrate their shields. Given that warp strafing is less intense an attack method than impulse combat, fighting it out in the way the Alliance fleet did may have seemed the only option besides withdrawal.
We've also seen occasions where a vessel has lost warp drive, too numerous to mention here. None are known that would specifically challenge what we've pondered so far, though any curious examples found or sent in will be appended here.
In short, there is nothing to suggest that warp strafing is impossible, and the fact that we have actually seen it with both phasers and torpedoes (not to mention knowing it can occur with transporters!) is the nail in the coffin of the idea that it can't be done.
The maneuver is of limited utility against high-tech targets that are warp-impaired or fixed in place, since its defensive advantages for the warp-strafing attacker require extreme velocity. This compounds the problems of targeting difficulty and low attack intensity, rendering warp strafing a less attractive option in most 23rd and 24th Century battle contexts.
However, against less advanced targets or those without significant FTL sensor capability, warp strafing could easily serve as an almost magical "death from nowhere" attack method.
Special thanks to Graham Kennedy for the "Let That Be Your Last Battlefield" info. Thanks to "The Darkling" for pointing out the "Future Tense" example, et cetera. And, of course, thanks to Trek5.